Myeongjang Pyeon

Output Video

0: fully recover the input / 1.05: fully rerender the input

Uniformly sample the key frames every K frames

Describe what you want to see in the output video

Zero-Shot Text-Guided Video-to-Video Translation